Auto-Registering Nested Tables in Unstructured Forms
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Automatic registration of forms with irregular or floating keyword and content layouts is challenging, as existing rules fail to accurately associate keywords with content in complex forms, leading to errors and difficulties in modifying rules for different formats.
Innovation Solution
A deep learning-based system that extracts features from forms, identifies tables, groups cells, forms hierarchies, and associates keywords with content using geometric and semantic distances, enabling auto-registration and handling nested keywords and content in a hierarchical data structure.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If fixed rules are used for keyword association, then simple forms can be processed easily, but complex forms with irregular layouts fail to associate keywords with content accurately
Solution Approach 1:
The patent segments the form processing task into multiple stages: table structure identification, cell grouping, hierarchical relationship establishment, and keyword-content association. This segmentation allows the system to handle complex irregular forms by breaking them down into manageable components while maintaining association accuracy.
Solution Approach 2:
The patent introduces a hierarchical dimension to the association process by establishing parent-child relationships between keywords and content based on table structure. This dimensional approach allows the system to associate keywords with content even when they are not in simple left-to-right or top-to-bottom positions, resolving the contradiction between ease of implementation and association accuracy.
2Adaptability or versatility
If modified rules are created to handle floating forms, then irregular layouts can be processed, but existing fixed forms may break and require re-testing
Solution Approach 1:
The patent creates a universal table-structure-based association algorithm that works for both fixed and floating forms. By using table structure as the foundation for all association rules, the system achieves layout adaptability while maintaining stability, as the same core algorithm handles different form types without requiring separate rule sets or re-testing.
Solution Approach 2:
The patent changes the fundamental parameter from position-based association rules to structure-based association rules. This parameter change allows the system to adapt to different layouts by interpreting table structure hierarchies rather than relying on fixed positional relationships, achieving both adaptability and stability simultaneously.
3Adaptability or versatility
If multiple tabular formats are used within a single form, then complex information can be represented, but keyword-content association becomes more difficult
Solution Approach 1:
The patent segments the form into multiple table structures and processes each table's hierarchy independently. By identifying table boundaries and establishing separate hierarchical relationships for each table, the system can handle multiple tabular formats within a single form without increasing overall processing complexity, as each table is processed using the same standardized approach.
Data Source
AI summary
In some forms containing keywords and content, there may be nested levels of keywords, also referred to as a hierarchy. Content in the forms may be associated with one or more keywords in one or more of the nested levels, or in the hierarchy. Identifying keywords in adjacent cells in a table (with a nested keyword being either to the right of or below another keyword) enables distinguishing between keywords and content in filled forms, and enables correct association of content with respective keywords.